PMDG descent woes (737-800)

[Mxsmanic]

Has anyone flown this route from KLAS to KSAN:

BOACH3.TNP.LYNDI2

Departing 25R landing RNAV 27.

I have tried several times to get my PMDG 737-800 (FS9) to descend gracefully
to landing after top of descent, but it just won't do it. Either it descends
too slowly, or I get errors like DES PATH UNACHIEVABLE. I've tried descending
with spoilers, choosing a slower descent speed, varying loads (from empty to
fully loaded), etc., and it's still not behaving.

I'm a bit rusty on the 737-800, but I don't think I'm forgetting anything. I'm
wondering if anyone else has the same troubles. The 737-800 has never been
easy to keep in a graceful descent, but this seems worse than I remember.

[Tom P]

I assume you're flying an approach managed by the FMC using VNAV?
Have a look at the LEGS pages and see what the speed/altitude profile is
for the various waypoints and check it against the published approach at
http://www.airnav.com/airport/ksan

You can also find the procedure in the PMDG folder as KSAN.txt. My copy
has MOMAR at or below 14000. Which part of the flight are you getting
the warning? If it's between TNP and MOMAR then you're probably too high
passing TNP.

Personally I don't think much of letting the autopilot manage the
vertical profile completely. Use your own judgement to decide the TOD
and only use autopilot to control the descent rate in VNAV.

[Mxsmanic]

Tom P writes:

> I assume you're flying an approach managed by the FMC using VNAV?

Right. But the FMC fails to command a descent rate adequate to make the
altitude restrictions and/or kicks out VNAV with DES PATH UNACHIEVABLE.

I also get a warning sometimes about STEEP DES AFTER TRIXI.

In the past I've found that reducing the commanded airspeed for the descent
phase in the FMC seems to help it to stick to the descent profile, but that
doesn't seem to be helping here.

> Have a look at the LEGS pages and see what the speed/altitude profile is
> for the various waypoints and check it against the published approach at
> http://www.airnav.com/airport/ksan

They match. TRIXI above 12000 and below 14000, SLOYD (11 miles beyond) above
10000.

At 270 knots, going from 14000 to 10000 in 11 miles requires about 1600 fpm,
which does not seem excessive. I'm not sure what the FMC sees that it
considers steep.

> You can also find the procedure in the PMDG folder as KSAN.txt. My copy
> has MOMAR at or below 14000.

The current LYNDI2 doesn't have a constraint and MOMAR, and the PMDG database
matches this.

> Which part of the flight are you getting the warning? If it's between
> TNP and MOMAR then you're probably too high passing TNP.

The FMC starts a descent but the descent rate is inexplicably gentle, then
later on it says DES PATH UNACHIEVABLE. With a fully-loaded aircraft, the
message takes longer to appear (and the FMC complains about a steep descent
after TRIXI), but it's still there, and the commanded descent rate is still
too gentle (I rapidly end up thousands of feet above the descent path).

I can make the descent myself in V/S mode, but the FMC seems to have trouble
with it.

> Personally I don't think much of letting the autopilot manage the
> vertical profile completely. Use your own judgement to decide the TOD
> and only use autopilot to control the descent rate in VNAV.

I'm mainly wondering if this behavior mirrors the real-world FMC. I'm trying
to do it as closely to real life as possible. I know some FMCs have trouble in
some situations in the real world, but I don't know if this is one of them or
if the PMDG implementation simply has a bug.

[Mxsmanic]

Hmm, I tried the same route in the PMDG 747-400 (FS9), and I had no problems.
The 737-800 FMC has always been less stable than the 747 FMC, I think. Still,
maybe I'm doing something wrong.

[Iain Smith]

It's always seemed to me that the TOD in the Boeing FMCs as modelled by
PMDG is far too late in the plan. I think anyone would struggle to get
the 738 down within normal passenger carrying limits of descent rate.

I always fly using Radar Contact and that descends me way sooner than
the FMC shows.

Iain
Rugby, UK

[Mxsmanic]

Iain Smith writes:

> It's always seemed to me that the TOD in the Boeing FMCs as modelled by
> PMDG is far too late in the plan. I think anyone would struggle to get
> the 738 down within normal passenger carrying limits of descent rate.

I wonder if there's something in the FMC that I can adjust to get an earlier
TOD.

> I always fly using Radar Contact and that descends me way sooner than
> the FMC shows.

I normally fly on VATSIM, so I descend when told, but these were practice
flights carried out offline or when ATC was not online.